Parameter Identification of Film Using Image Processing : 1st Report, Fundamental Study

Abstract

In this paper, we intend to identify the density of a small element of a film. The interference fringes on the film were measured by a CCD camera and were processed to estimate the deflection of the film. Then, the density distribution was identified. Because the gray level quantization and the resolution of the CCD degraded the accuracy of the estimated deflection and the identified density, we reduced the noise in the image data by the Wavelet denoising technique. Also, since the identification process was very sensitive to the noise, we applied the Tikhonov regularization technique to regulate the identification process. Finally, the Tikhonov regularization technique gave the precise density distribution by using well denoised image data.